arXiv cs.LG - 2026-08-14

268 items collected.

1. LoKiFormer: Locality-aware Attention with Decoupled Knowledge Memory for Efficient Large Language Model Pretraining

Author: Qiuwu Chen, Zimo Liu, Yuchen Li, Ying Sun, Yifan Zhang, Zhijie Qiu, Zeng You, Ryan Dong, Simeng Ma, Yaofo Chen, Mingkui Tan
Published: 8/14/2026, 4:00:00 AM
Categories: cs.LG

arXiv:2608.12419v1 Announce Type: new Abstract: Large language models (LLMs) have achieved remarkable breakthroughs across various applications. However, their architectures remain inefficient in pretraining due to two main limitations: (i) self-attention lacks an explicit inductive bias for localit...

arXiv cs.LG - 2026-08-14 ​

1. LoKiFormer: Locality-aware Attention with Decoupled Knowledge Memory for Efficient Large Language Model Pretraining ​

2. Which Site, and When: A Free-Satellite-Data Test of Himalayan Glacial Lake Bursts, Landslides, and Ice Floods ​

3. MARCH: Scaling Recurrent Memory with Content-Routed State Anchors ​

4. Multi-AUV Ad-hoc network-based Target Tracking: A Value Gradient Guidance Multi-Agent Diffusion Reinforcement Learning Approach ​

5. Unifying Generative Models with Path Integrals ​

6. Dual Spatial-Temporal Attribution: Architecture-Aligned Post-Hoc Explainability for Recurrent Graph Anomaly Detection ​

7. Personalized Scorer Modeling: A Learning-Based Framework for Deriving Robust Sleep Stage Labels from Multiple Experts ​

8. Geometric and Behavioral Stratification in Transformer Residual Streams ​

9. Exemplar-based objective classification of gust-induced loads across multiple flight conditions ​

10. Learning Under Treatment-Induced Label Indeterminacy with Expert Annotations of Counterfactual Outcomes: A Case Study in Neurological Prognostication ​

11. When Can You Trust Offline Evaluation of Equal-Cost Top-k Allocation? A Controlled, Reproducible Benchmark and Practitioner's Guide ​

12. Exploring Oversmoothing with Householder Matrices ​

13. GENADA: efficient generative time series adversarial attack framework ​

14. Scaling Automatic Research Agents via World Models ​

15. Prof-K: Probabilistic One-Pass Filtering for Efficient Top-k Selection ​

16. Represent, Then Generate: Multimodal-Conditioned Time-Series Generation under Irregular Missingness ​

17. Predicting When Random Low-Dimensional Reparameterizations Train Neural Networks ​

18. The Boolean Power of ReLU ​

19. Structure-preserving uncertainty quantification for GENERIC dynamics ​

20. CAKE: Compiler-Agent Co-Design for Frontier Kernel Evolution ​

21. Interpretable Causal Discovery via Causal-Effect Constraints ​

22. Training Under Challenge: Executable Certificates and Challenge-Closed Optimality for Neural Networks ​

23. Demand Transfer Estimation at Scale via Restricted Logit Modeling ​

24. Finding the Needle in a Haystack: Test-Time Analog Circuit Representation Adaptation for Bayesian Optimization ​

25. The Impact of Temporal Context Length and Encoding Strategies on Self-Supervised ECG Representation Learning ​

26. A Contract-Grade Verifier for LLM-Generated GPU Kernels, and a Native Blackwell Backward for the Gated-Linear-Recurrence Family ​

27. Federated Compositional Muon Optimizer for Matrix-Wise Models ​

28. Perturbation-based Regional Interpretability through Subtraction Mapping (PRISM): naming-error dissociations in language models and post-stroke aphasia ​

29. MAG: MAnifold Guided Semi-Supervised Multi-modal In-Context Learning ​

30. A Cloud-Edge System for Multimodal Clinical Screening in Resource-Constrained Rural Settings ​

31. Decentralized Multi-Player Q-Learning in Episodic Markov Decision Processes with Information Asymmetry ​

32. Beyond Outcome Rewards: Step-Level Self-Distilled Policy Optimization for Deep Search Agents ​

33. CoMedBench: A Multi-Source Benchmark of Synthetic Medical Data Fidelity and Downstream Utility ​

34. HiRoute: Hierarchical Routed Prompt Tuning for Safety Alignment of Large Language Models ​

35. Fast A/B/n Testing: Exact Multi-Policy Comparison via Tree-Coupled Feedback Sharing ​

36. A Compositional Theory of Curvature in Probabilistic Circuits ​

37. Sustaining Plasticity via Learnable Wavelet Activations in Continual Learning ​

38. Robust data-driven discovery of fractional differential equations via weak formulations and Pareto-based subset selection ​

arXiv cs.LG - 2026-08-14

1. LoKiFormer: Locality-aware Attention with Decoupled Knowledge Memory for Efficient Large Language Model Pretraining

2. Which Site, and When: A Free-Satellite-Data Test of Himalayan Glacial Lake Bursts, Landslides, and Ice Floods

3. MARCH: Scaling Recurrent Memory with Content-Routed State Anchors

4. Multi-AUV Ad-hoc network-based Target Tracking: A Value Gradient Guidance Multi-Agent Diffusion Reinforcement Learning Approach

5. Unifying Generative Models with Path Integrals

6. Dual Spatial-Temporal Attribution: Architecture-Aligned Post-Hoc Explainability for Recurrent Graph Anomaly Detection

7. Personalized Scorer Modeling: A Learning-Based Framework for Deriving Robust Sleep Stage Labels from Multiple Experts

8. Geometric and Behavioral Stratification in Transformer Residual Streams

9. Exemplar-based objective classification of gust-induced loads across multiple flight conditions

10. Learning Under Treatment-Induced Label Indeterminacy with Expert Annotations of Counterfactual Outcomes: A Case Study in Neurological Prognostication

11. When Can You Trust Offline Evaluation of Equal-Cost Top-k Allocation? A Controlled, Reproducible Benchmark and Practitioner's Guide

12. Exploring Oversmoothing with Householder Matrices

13. GENADA: efficient generative time series adversarial attack framework

14. Scaling Automatic Research Agents via World Models

15. Prof-K: Probabilistic One-Pass Filtering for Efficient Top-k Selection

16. Represent, Then Generate: Multimodal-Conditioned Time-Series Generation under Irregular Missingness

17. Predicting When Random Low-Dimensional Reparameterizations Train Neural Networks

18. The Boolean Power of ReLU

19. Structure-preserving uncertainty quantification for GENERIC dynamics

20. CAKE: Compiler-Agent Co-Design for Frontier Kernel Evolution

21. Interpretable Causal Discovery via Causal-Effect Constraints

22. Training Under Challenge: Executable Certificates and Challenge-Closed Optimality for Neural Networks

23. Demand Transfer Estimation at Scale via Restricted Logit Modeling

24. Finding the Needle in a Haystack: Test-Time Analog Circuit Representation Adaptation for Bayesian Optimization

25. The Impact of Temporal Context Length and Encoding Strategies on Self-Supervised ECG Representation Learning

26. A Contract-Grade Verifier for LLM-Generated GPU Kernels, and a Native Blackwell Backward for the Gated-Linear-Recurrence Family

27. Federated Compositional Muon Optimizer for Matrix-Wise Models

28. Perturbation-based Regional Interpretability through Subtraction Mapping (PRISM): naming-error dissociations in language models and post-stroke aphasia

29. MAG: MAnifold Guided Semi-Supervised Multi-modal In-Context Learning

30. A Cloud-Edge System for Multimodal Clinical Screening in Resource-Constrained Rural Settings

31. Decentralized Multi-Player Q-Learning in Episodic Markov Decision Processes with Information Asymmetry

32. Beyond Outcome Rewards: Step-Level Self-Distilled Policy Optimization for Deep Search Agents

33. CoMedBench: A Multi-Source Benchmark of Synthetic Medical Data Fidelity and Downstream Utility

34. HiRoute: Hierarchical Routed Prompt Tuning for Safety Alignment of Large Language Models

35. Fast A/B/n Testing: Exact Multi-Policy Comparison via Tree-Coupled Feedback Sharing

36. A Compositional Theory of Curvature in Probabilistic Circuits

37. Sustaining Plasticity via Learnable Wavelet Activations in Continual Learning

38. Robust data-driven discovery of fractional differential equations via weak formulations and Pareto-based subset selection